Putting open source arguments to one side, normal economics says that if a well-healed monopoly with a significant investment entry barrier required to enter it, then that market is pretty much unassailable; very few businesses would risk trying to take it- the encumbent can too readily enter a price cutting war of attrition.
The only way an *unearned* monopoly can exist is through government force.
Depends what you mean by "earned". If you mean is it impossible for a rich, but otherwise clueless company to buy up all the competitors in any sector and achieve monopoly status- then no, of course not. And there are other ways this can come about, for example if a major competitor folds.
It's a massive mistake, because the evidence that increased piracy actually doesn't seem to affect sales at all, if at all, and may even improve sales.
It's not clear why this would be, but there are suggestions that essentially there has been piracy ever since home taping in general and cassettes in particular hit the market- people have been copying off of friends ever since.
Now, you could argue that this is wrong, and that the artists and music companies are worse off because of this. But music companies are making good money, and stopping the home copying would be a double-edged sword, since the home copying acts as free advertising. In addition, it's very unclear that the lack of home copying would increase sales- many people, particularly young people are on a budget, and simply wouldn't buy more music, they would just listen to less music and spend no more money.
Also, except for the most hardened copier listening to music usually creates a taste for music- so they end up buying more music in the long run.
So, home copying doesn't seem to reduce the market size for selling music. On the other hand, real pirates- people making copies of music and selling them for money, or even as legitimate forgeries, they really can reduce the market size.
So, "Dumbass", you're saying that if I'm behind a NAT, and you over there are behind a NAT, Skype is going to establish a connection to you through me?
Wrong. LOL. I'm saying it gets routed through a third party.
The whole point of "through me" connections is that the NATed box and the recipient box are on the same internal network.
Wrong.
Your tone of voice and lack of understanding of networks is highly indicative that you are a freshman that thinks they got the whole world figured out.
Wrong, wrong, wrong! ROFLMAO.
Seriously, just read section 4.5 of the document you've evidently never, ever read and quit making an ass of yourself.
My point- dumbass- is that if they are behind *different* NAT firewalls (actually routers). Then they specifically *do* send messages from A to B to C; something that you said "makes no sense".
And the protocol has to do that because NATd boxes can't ordinarily accept incoming connections (atleast not without the administrator doing lots of prior manual set-up on the NAT router). So point B has to be chosen from the set of non NAT'd IP addresses.
In most cases the voice packets go direct leaf node to leaf node.
However if both are behind NAT firewalls then they can't directly talk to each other, and the Skype protocol seems to pick another Skype users machine (picked by some scheme that probably isn't publically described anywhere) and route the packets through them.
There's no security problem with doing this (the packets are end-end encrypted), it just takes longer and is more likely to congest, and it takes up bandwidth at the extra users network.
Really, IMNHO, NAT is a real menace. I'm really looking forward to IPv6 which doesn't have all this garbage; should be here by 2100 I reckon.
Psuedoscience is inherently not reproducible in carefully controlled, well designed experiments. This certainly seems to be reproducible, hence is not psuedoscience.
I think you're assuming that the design is the absolute best possible that could be built at the time.
In engineering, a good-rather than best, easily tested, off-the-shelf approach is more typically employed. This isn't necessarily the cheapest, or the best, or the most reliable. But it is easily defended in design reviews.
The elevator is positioned by a combination of centrifugal force acting directly outwards from the rotation axis of the earth, and the earths gravity.
It turns out that if you add the two sets of forces together you get a force that pushes the cable towards the plane above the equator as well as the force acting away/towards the earth.
It's easiest to see at the poles. If you imagine swinging a long cable around the north pole the rotation pulls it out in a long line, but the earths gravity would pull it down towards the equator. It would probably drag along the ground in fact; although if you built a tall tower you might be able to stop that.
Whilst you can fix the lower end of the cable somewhat north or south of the equator, the geosynchronous orbit part of the cable ends up pretty much in the equatorial plane (it will move a bit as you move the fixture point, but not much, since the cable is fairly flexible.)
If the fixture point goes too far north or south, the cable will actually drag along the ground.
If you stick a massive counterweight on the cable then the cable becomes taut and can be pulled off the ground somewhat, but you're ultimately limited by the strength/weight of the cable- making the cable tauter needs it to be stronger, which makes it heavier. So there's a limit how far north/south you can go in practice.
Oh well, if they're "dated" then they must be bad. Sure hope your girlfriend has only gone out with you in her entire life.:-)
Seriously though, the reason rockets are expensive is because they aren't launched very much- mass production would slash the cost. But because the cost is high, production is low, and so nobody can afford to go, so the cost stays high.
If that sounds utopian, consider that the fuel to put somebody into orbit is only about as much as to send someone on a round the world trip by jet...
Rocket hardware, contrary to popular opinion, isn't very complicated, your car probably is about as complicated.
Incidentally, the projected cost of Space Elevators is likely to be about as high as rockets- it's only if the launch rate goes really high will the initial higher R&D costs of Space Elevators cancel out.
Then there's the Van Allen radiation belts around the earth- people would get radiation sickness and possibly die if they go up an elevator. Shielding is extremely heavy and expensive, but rockets go much faster so you get less dossage and rockets can do what Apollo did, steer around the worst of the belts- but elevators have to be above the equator where the belts are, so they can't do that.
Even then, there's another fly in the ointment, the power costs of a space elevator are much higher than you would expect- currently the costs per kg to orbit are thought to be higher than the cost of cheap rocket fuel to do the same thing. This is mainly because the laser power beaming system looks like it may turn out to be about 2% efficient for various reasons (and even that's optimistic- current tech is 0.5% efficient), and other techniques aren't practical for sending power 38000 km up a nanotube rope. It turns out that rockets are if anything more efficient, and may even be cheaper in the long term.:-(
[or:-) if you like rockets, personally I like all ways to get to space:-) ]
Also, since News services, as far as we know, isn't DNA based, doesn't use inheritance and wasn't bred using natural selection- I don't really think that Darwin would exactly be an authority you would want to ask even if he was extremely young for his age.
Oh, yeah, I forgot- there's one annoying difference between Java and C/C++.
In Java if you modify the code of in a class method, if everything works perfectly [which it never seems to:-) ] it triggers recompilation of all the dependent classes- even though they don't actually need recompilation.
In C/C++, modifying the.c or.cc file only triggers recompilation of the.o file and relinking.
That can make an *enormous* difference on a big system- and modification of code without modifying the header file is very common. On a mature system that can save >>50% of compiling.
I don't think *any* of the Java toolsets deal with that issue.
If the Java system wrote out the 'header' separately from the code then before replacing the header file after recompiling the.java it could check to see if it had really changed, if it hadn't it could leave it alone; and that would preclude it doing unnecessary work.
Just watching the files opened is dangerous because Java can generate multiple.class file per source because of inner classe (example: a.java can generate a.class, a$1.class, a$2.class, etc). The tool may miss something.
Actually, no, because in order to write the files out the compiler must have opened them. The clearcase technique is very clever and general. Still, Ant seems pretty good from what you say.
Yeah but Gcc for example spits out the dependency information if you ask it to, and then you can suck that into a makefile automatically. Then the makefile only needs to check the timestamp on the files to know whether it needs to rebuild everything and nothing that doesn't. I think I benchmarked that at 10,000 files in 45 seconds plus compilation time.
Java's compiling technique seems to me to be quite broken- only recompiling the.java files that have changed often results in runtime problems. If you're lucky it will be an error otherwise it will just do something strange. I'm not sure whether ant has a fix for that; it's theoretically possible, but I don't know it well enough; I doubt it. I think the clearcase tool can do this properly with Java, because it watches the files that the compiler opens. But the last place I worked didn't set it up properly and they just recompiled everything every time.
One thing I've noticed is that the automatic recompilation feature of Java seems broken- changing a file doesn't reliably trigger recompilation of all affected.class files. That's really bad for big projects since you basically have to recompile everything from scratch.
I've never quite got to the bottom of how and why, but I beleive that if Java had automatically generated separate header files and code files and dependency lists from the.java files then it probably would have worked correctly, since the compile could check the date on the header file and only recompile code files that need recompilation- much like C does (on a good day with decent tools, which alas are rare).
Slashdot Mirror
The only way an *unearned* monopoly can exist is through government force.
Depends what you mean by "earned". If you mean is it impossible for a rich, but otherwise clueless company to buy up all the competitors in any sector and achieve monopoly status- then no, of course not. And there are other ways this can come about, for example if a major competitor folds.
It's not clear why this would be, but there are suggestions that essentially there has been piracy ever since home taping in general and cassettes in particular hit the market- people have been copying off of friends ever since.
Now, you could argue that this is wrong, and that the artists and music companies are worse off because of this. But music companies are making good money, and stopping the home copying would be a double-edged sword, since the home copying acts as free advertising. In addition, it's very unclear that the lack of home copying would increase sales- many people, particularly young people are on a budget, and simply wouldn't buy more music, they would just listen to less music and spend no more money.
Also, except for the most hardened copier listening to music usually creates a taste for music- so they end up buying more music in the long run.
So, home copying doesn't seem to reduce the market size for selling music. On the other hand, real pirates- people making copies of music and selling them for money, or even as legitimate forgeries, they really can reduce the market size.
It's called deafness; often with a side order of tinnitus. :-(
Or vice-versa :-)
Seriously man- change the password to kjhs45ks, it's just not worth the risk to your Karma.
Wrong. LOL. I'm saying it gets routed through a third party.
The whole point of "through me" connections is that the NATed box and the recipient box are on the same internal network.
Wrong.
Your tone of voice and lack of understanding of networks is highly indicative that you are a freshman that thinks they got the whole world figured out.
Wrong, wrong, wrong! ROFLMAO.
Seriously, just read section 4.5 of the document you've evidently never, ever read and quit making an ass of yourself.
My point- dumbass- is that if they are behind *different* NAT firewalls (actually routers). Then they specifically *do* send messages from A to B to C; something that you said "makes no sense".
And the protocol has to do that because NATd boxes can't ordinarily accept incoming connections (atleast not without the administrator doing lots of prior manual set-up on the NAT router). So point B has to be chosen from the set of non NAT'd IP addresses.
Um, actually RTFA.
In most cases the voice packets go direct leaf node to leaf node.
However if both are behind NAT firewalls then they can't directly talk to each other, and the Skype protocol seems to pick another Skype users machine (picked by some scheme that probably isn't publically described anywhere) and route the packets through them.
There's no security problem with doing this (the packets are end-end encrypted), it just takes longer and is more likely to congest, and it takes up bandwidth at the extra users network.
Really, IMNHO, NAT is a real menace. I'm really looking forward to IPv6 which doesn't have all this garbage; should be here by 2100 I reckon.
Psuedoscience is inherently not reproducible in carefully controlled, well designed experiments. This certainly seems to be reproducible, hence is not psuedoscience.
You obviously have never driven a Ford :-)
In engineering, a good-rather than best, easily tested, off-the-shelf approach is more typically employed. This isn't necessarily the cheapest, or the best, or the most reliable. But it is easily defended in design reviews.
They probably were available, they'd just have been very expensive.
No biggie, you just have to make the bolus section long enough- like 1000km long enough :-)
The difficult/expensive bits are building something 200km high, and having the ends of the string being at orbital velocity :-)
But it is theoretically possible, particularly if nanotubes are used for the 'string'.
That's not actually so. The Russian Proton rocket engines for example, uses a safety factor of 2.
i.e. they worked out the stress on each of the parts and then made them twice as strong as they needed to.
It turns out that if you add the two sets of forces together you get a force that pushes the cable towards the plane above the equator as well as the force acting away/towards the earth.
It's easiest to see at the poles. If you imagine swinging a long cable around the north pole the rotation pulls it out in a long line, but the earths gravity would pull it down towards the equator. It would probably drag along the ground in fact; although if you built a tall tower you might be able to stop that.
Whilst you can fix the lower end of the cable somewhat north or south of the equator, the geosynchronous orbit part of the cable ends up pretty much in the equatorial plane (it will move a bit as you move the fixture point, but not much, since the cable is fairly flexible.)
If the fixture point goes too far north or south, the cable will actually drag along the ground.
If you stick a massive counterweight on the cable then the cable becomes taut and can be pulled off the ground somewhat, but you're ultimately limited by the strength/weight of the cable- making the cable tauter needs it to be stronger, which makes it heavier. So there's a limit how far north/south you can go in practice.
Seriously though, the reason rockets are expensive is because they aren't launched very much- mass production would slash the cost. But because the cost is high, production is low, and so nobody can afford to go, so the cost stays high.
If that sounds utopian, consider that the fuel to put somebody into orbit is only about as much as to send someone on a round the world trip by jet...
Rocket hardware, contrary to popular opinion, isn't very complicated, your car probably is about as complicated.
Incidentally, the projected cost of Space Elevators is likely to be about as high as rockets- it's only if the launch rate goes really high will the initial higher R&D costs of Space Elevators cancel out.
Then there's the Van Allen radiation belts around the earth- people would get radiation sickness and possibly die if they go up an elevator. Shielding is extremely heavy and expensive, but rockets go much faster so you get less dossage and rockets can do what Apollo did, steer around the worst of the belts- but elevators have to be above the equator where the belts are, so they can't do that.
Even then, there's another fly in the ointment, the power costs of a space elevator are much higher than you would expect- currently the costs per kg to orbit are thought to be higher than the cost of cheap rocket fuel to do the same thing. This is mainly because the laser power beaming system looks like it may turn out to be about 2% efficient for various reasons (and even that's optimistic- current tech is 0.5% efficient), and other techniques aren't practical for sending power 38000 km up a nanotube rope. It turns out that rockets are if anything more efficient, and may even be cheaper in the long term. :-(
[or :-) if you like rockets, personally I like all ways to get to space :-) ]
IRC Richard Dawkins was rather against the use and essentially overuse of that particular metaphor.
Also, since News services, as far as we know, isn't DNA based, doesn't use inheritance and wasn't bred using natural selection- I don't really think that Darwin would exactly be an authority you would want to ask even if he was extremely young for his age.
In Java if you modify the code of in a class method, if everything works perfectly [which it never seems to :-) ] it triggers recompilation of all the dependent classes- even though they don't actually need recompilation.
In C/C++, modifying the .c or .cc file only triggers recompilation of the .o file and relinking.
That can make an *enormous* difference on a big system- and modification of code without modifying the header file is very common. On a mature system that can save >>50% of compiling.
I don't think *any* of the Java toolsets deal with that issue.
If the Java system wrote out the 'header' separately from the code then before replacing the header file after recompiling the .java it could check to see if it had really changed, if it hadn't it could leave it alone; and that would preclude it doing unnecessary work.
Officers will be along later to deport you to Guantanamo bay, since you have clearly had interaction with terrorists.
It's probably more to do with Harlow. I mean, have you *been* to Harlow? :-)
It can't be! Posh and Becks live near there, and even more impressively so do *I* :-)
Actually, no, because in order to write the files out the compiler must have opened them. The clearcase technique is very clever and general. Still, Ant seems pretty good from what you say.
Java's compiling technique seems to me to be quite broken- only recompiling the .java files that have changed often results in runtime problems. If you're lucky it will be an error otherwise it will just do something strange. I'm not sure whether ant has a fix for that; it's theoretically possible, but I don't know it well enough; I doubt it. I think the clearcase tool can do this properly with Java, because it watches the files that the compiler opens. But the last place I worked didn't set it up properly and they just recompiled everything every time.
One thing I've noticed is that the automatic recompilation feature of Java seems broken- changing a file doesn't reliably trigger recompilation of all affected .class files. That's really bad for big projects since you basically have to recompile everything from scratch.
I've never quite got to the bottom of how and why, but I beleive that if Java had automatically generated separate header files and code files and dependency lists from the .java files then it probably would have worked correctly, since the compile could check the date on the header file and only recompile code files that need recompilation- much like C does (on a good day with decent tools, which alas are rare).